Testing the Old Adage: Does Defense Win Championships?

Question

Testing the old saying, does “defense win championships”? I analyze Super Bowl performances from 1967-2020, comparing historically great teams - ones that have won more than two Super Bowls - to teams that have had brief success. Is having a superior defense the key to sustained Super Bowl success? If not, what is?

Data Collection

This Super Bowl data can be credited to Timo Bozsolik, who posted the data on Kaggle. This data contains Super Bowl data from Super Bowl I in 1967, all the way until Super Bowl LIV in 2020. It contains data on the winner, loser, points scored by each team, MVP, and location of the game. I mutate point differential to the dataset, and quickly add Super Bowl LV data from 2021.

The draft grades data is taken from the following Bleacher Report article, and I have converted the letter grades (A-F) to numeric values of 1-13, with 1 being an A+, 2 being an A, all the way to F, which is 13. I also created a column in the excel file to represent the number of Super Bowl appearances for each organization from 2003-2021. I chose the given time frame for Super Bowl appearances since the Bleacher Report article provides data from 2003-2012 and drafts have a long-term effect as well. For example, players from a draft in 2012 may develop a few years down the road and maximize their impact in 2017, although this is a less certain possibility.

Analysis

Super Bowls are generally tight affairs, as shown in the graph below. Although there is a large range of point differential, Super Bowls have been decided by a median point differential of just above 10 points since 1967, as represented by red dot on the animated plot. Especially as the median point differential has decreased since the turn of the century, when games are this tight, it makes sense that a superior defense would make this marginal difference between losing and winning a championship.

Let’s dive in.

Sustained, or historically great, winners allow a median of 17 points in the Super Bowl, while those with limited success allow a median of 10 points. Although historically great teams have a minimum lower points allowed, it is clear to see that, surprisingly, sustainably great teams don’t necessarily have superior defenses, at least based on Super Bowl performances. This doesn’t say anything about the importance of defense in the regular season, but winning more Super Bowls isn’t associated with superior defensive performances.

Historically great teams score an even lower median of 29 points in the Super Bowl, compared to a median of 31 points for non-historically great teams. However, the range for points scored is greater for historically great teams. It’s interesting that the organizations with non-sustained success appear to have both offenses and defenses that perform better in the Super Bowl, quite contrary to what I initially believed.

The aforementioned notion is validated by this graph, with historically great organizations playing in closer Super Bowls, as the median point differential is 10 points. This is much less than the median point differential of 16 points for organizations with non-sustained success. Although the difference between the two median point differentials may appear marginal, 6 points is the equivalent of one touchdown. This means that those with non-sustained success win Super Bowls by a median differential of almost 3 scores, compared to a much tighter 2-score differential for sustainable winners.

Why is this the case? Let’s examine the figure above, which depicts the relationship between Super Bowl appearances from 2003-2021 and Draft Grades for teams between 2003-2012.

My findings so far lead me to hypothesize that organizations with brief success go all out for those 1 or 2 particular years, occasionally capitalizing on high draft picks and free agency signings. However, their lack of sustained success may come down to a lack of long-term planning, such as a steady stream of poor drafts and free agency signings that don’t provide any long-term benefit. They may get things right for a season or two, but for the most part, they fail.

This largely holds true based on the figure above, as organizations like the New England Patriots, Pittsburgh Steelers, and Seattle Seahawks, which have had more than two Super Bowl appearances since 2003, have grades of A, A+, and B-, respectively. While there are teams like the Cowboys and Texans that have B grades but haven’t made the Super Bowl, only teams with a C draft grade or higher have made the Super Bowl since 2003. Furthermore, teams like the Packers and Ravens, which have had only one appearance since 2003 but have high draft grades, enjoy plenty of playoff appearances and success without making the championship. There definitely is a relationship between quality of draft picks and long-term success.

Conclusion

The NFL is full of talent, but talent can come and go in the league, either due to money, injury, or retirement. In order to create sustained success, organizations need to effectively plan for the future, either through management, free agency transactions, or draft picks.